Distributed Inference Processing Across Expansion Apparatuses
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing inference apparatuses face resource insufficiency when performing complex processing, as they cannot effectively utilize the hardware resources of connected expansion devices for inference processing.
Innovation Solution
The solution involves generating programs for distributing inference processing between the inference apparatus and connected expansion apparatuses, based on their respective hardware capabilities, allowing for efficient allocation of arithmetic processes across both devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If inference processing is performed using only the inference apparatus, then the system structure is simple, but the hardware resources cannot handle complex processing tasks
Solution Approach 1:
The inference processing is divided into two segments: the first program executes on the inference apparatus and the second program executes on the connected expansion apparatus. This segmentation allows the system to handle complex processing tasks by distributing workloads across multiple devices while maintaining a relatively simple individual device structure.
2Productivity
If an expansion apparatus is connected to the inference apparatus, then the processing capacity is improved, but the hardware resources of the inference apparatus itself cannot be utilized for inference processing
Solution Approach 1:
The system merges the inference apparatus and expansion apparatus into a unified inference processing system. The control unit generates a combined program that distributes inference tasks across both devices, allowing hardware resources of both the inference apparatus and expansion apparatus to be utilized simultaneously for inference processing, thereby improving overall resource utilization efficiency.
3Measurement precision
If complex inference processing is performed, then the inference accuracy is improved, but the hardware resources of the inference apparatus become insufficient
Solution Approach 1:
The system transitions from a single-device architecture to a multi-device distributed architecture. By adding the dimension of spatial distribution across multiple hardware devices (inference apparatus and expansion apparatus), the system can execute complex inference models that require more hardware resources than a single device can provide, thereby achieving high inference accuracy without being constrained by single-device resource limitations.
Data Source
AI summary
An information processing apparatus generates a program for executing inference processing using a learned inference model, the generated program including a first program for executing first processing, the first program being generated based on first information concerning inference processing hardware of a first inference apparatus and a second program for executing second processing, the second program being generated based on second information concerning inference processing hardware of one or more second inference apparatus, and distributes the inference processing to the first inference apparatus to execute first processing and to one or more second inference apparatus connectable to the first inference apparatus to execute second processing.